Key words: formalin-fixed paraffin-embedded (FFPE) • GenomiPhi • whole-genome amplification (WGA) • degenerate oligonucleotideprimed PCR (DOP-PCR) • Cy3/Cy5-dCTP probe labeling • comparative genomic hybridization (CGH) • bacterial artificial chromosome (BAC) array • array CGH
Human genomic DNA isolated from formalin-fixed paraffin-embedded (FFPE) tissue samples was amplified using the GenomiPhi™ DNA Amplification Kit, fluorescently labeled with Cy™3 and Cy5 dyes, and hybridized to human BAC microarrays. Commercially available human BAC arrays from Spectral Genomics were chosen to verify the methods. Array data quality was similar to that obtained using the DOP-PCR method, demonstrating the application of GenomiPhi DNA Amplification Kit in array CGH experiments.
Formalin-fixed paraffin-embedded (FFPE) tissues have been widely used to archive samples obtained from biopsies of various human cancers and diseased tissues linked to genetic abnormalities. The characterization of FFPE samples by clinical laboratories is mostly limited to cytogenetic techniques that analyze genetic changes at the chromosomal level (1). Because DNA purified from FFPE tissue slides is highly degraded, damaged, and available only in small quantities, there has been little molecular analysis of these samples. Extracting molecular information on genome integrity and DNA copy-number changes from FFPE samples has the potential to drastically improve our understanding of genetic disorders underlying disease states.
With the emergence of comparative genomic hybridization (CGH) using microarrays spotted with cloned DNA sequences, such as bacterial artificial chromosome (BAC) arrays, it has become possible to obtain whole-genome information on allelic differences between samples from normal vs d Technology April/May 2003. [Online.] http://www.amershambiosciences.com/applic/upp00738.nsf/vLookupDoc/253872879-B500/$file/GPT_Mamone_Reprint.pdf
5. Jones, K. E. et al. DNA Isolation and amplification from formalin fixed, paraffin embedded tissues. Presented at American Society for Human Genetics Annual Meeting, Toronto, Canada, 2004.
6. Cardoso, J. et al. Genomic profiling by DNA amplification of laser capture microdissected tissues and array CGH. Nucleic Acids Res. 32, e146 (2004).
7. Wang, G. et al. DNA amplification method tolerant to sample degradation. Genome Res. 14, 2357–2366 (2004).
Nucleon HT Genomic DNA Extraction Kit RPN8509
back to top iseased tissues (2, 3). CGH experiments can be performed with DNA purified from cell culture or generated using amplification methods for fluorescent probe labeling, particularly when sample amount is limited.
Profiling DNA copy-number changes in FFPE DNA has the potential to reveal genetic information on particular regions of genomes that might be altered in clinically relevant samples. However, FFPE samples pose a unique challenge to CGH analysis because DNA isolated from them is degraded and requires amplification before array analysis.
Whole-genome amplification (WGA) methods are used to yield sufficient template for probe labeling and array hybridization. A common such method is degenerate oligonucleotide–primed PCR (DOP-PCR).
In contrast to DOP-PCR, GenomiPhi™ DNA Amplification Kit provides WGA using a simple isothermal method (4). This method employs the unique biochemical properties of Phi29 DNA polymerase, a highly processive enzyme with excellent strand displacement and proofreading ability, to amplify DNA.
In this application note, we show array CGH data on DNA isolated from human liver FFPE slides, comparing DOP-PCR with GenomiPhi DNA Amplification Kit. The amplification methods are validated by comparing microarray hybridization data on human BAC arrays. The results show that the GenomiPhi method provides robust whole-genome amplification of FFPE material, with array CGH data quality similar to the DOP-PCR method.
GenomiPhi DNA Amplification Kit, 25-6600-01
Typhoon™ 8600 Variable Mode Imager inquire
Proteinase K, 100 mg E76230Y
SpectralChip™ 2600 BAC Array, 5 pack SC2600-5PK*
Constitutional Chip™ 1.0 BAC Array, 5 pack CC1.0-5PK*
*GE Healthcare is the exclusive distributor of Spectral Genomics products outside the United States and Canada.
Other materials required
FFPE human liver tissue (Asterand)
Chelex-100 resin (Bio-Rad)
DOP-PCR primers (Integrated DNA Technologies)
GenePix™ 4000B Microarray Scanner (Axon Instruments)
GenePix Pro 3.1 Microarray Image Analysis Software (Axon Instruments)
Thermal cycler (MJ Research)
Quant-iT™ PicoGreen™ dsDNA Assay Kit (Invitrogen)
HyperLadder I DNA ladder (BioLine)
Dpn II restriction endonuclease (New England BioLabs)
DNA extraction from FFPE tissue slides
Genomic DNA was isolated using Chelex-100 from a 20-µm section of FFPE liver tissue. The detailed protocol is given below. DNA was quantitated using PicoGreen dsDNA quantitation reagent.
1. Add 2 µl of 10 mg/ml Proteinase K to each tube (final concentration 200 µg/ml).
2. Incubate at 55 °C for 3 h. Agitate gently every hour.
3. Add 100 µl of 5% Chelex-100 in Tris-EDTA.
4. Heat to 99 °C for 10 min.
5. Shake gently and spin at 10,500 x g for 15 min (while still hot).
6. Place on ice and remove any hard ened wax with pipette tip.
7. Heat sample to 45 °C briefly.
8. Add 100 µl chloroform.
9. Agitate gently and centrifuge at 10,500 x g for 15 min.
10. Remove top phase (~180 µl) to fresh tube.
Amplification by GenomiPhi DNA Amplification Kit
After DNA isolation, 10 ng was amplified using GenomiPhi DNA Amplification Kit. Amplification without a heat denaturation step (formalin fixation also contributes to DNA denaturation) was completed over 18 h at 30 °C, followed by enzyme inactivation by heating at 65 °C for 10 min. Yield was determined by the PicoGreen dsDNA method. Suitability for use in Phi29 DNA polymerase�mediated amplification was evaluated by monitoring the level of inhibition in the amplification reaction. Quality of genomic DNA and amplified product were determined by conventional PCR (5) and array CGH, respectively.
Amplification by DOP-PCR
Amplification of FFPE-extracted DNA by DOP-PCR was carried out using the Wellcome Trust Sanger Institute protocol (3). The detailed method can be obtained from http://www.sanger.ac.uk. Briefly, the protocol uses three random primers (DOP1, DOP2, and DOP3) for genomic DNA amplification using a thermophilic DNA polymerase.
Labeling and microarray hybridization
To prepare amplified DNA for microarray analysis, Cy3- and Cy5-labeled probes were generated from both amplification methods.
Cy3 and Cy5 probe labeling of amplified DNA was carried out as described in the SpectralChip 2600 manual (Spectral Genomics). This protocol is outlined in greater detail in reference 3. Genomic DNA (500 ng–2 µg) was labeled using the random priming method in the presence of Cy3- and Cy5-labeled dCTP and Klenow fragment. Cy3- and Cy5 labeled probes were separated on a 2% agarose gel for size analysis and imaged using a Typhoon 8600 Variable Mode Imager. Hybridization was performed on SpectralChip 2600 BAC Arrays according to the manufacturer’s recommended conditions. The arrays were scanned on a GenePix 4000B scanner and signal intensities quantitated using GenePix Pro 3.1 software.
DNA extraction from FFPE tissue
DNA was extracted from human liver FFPE tissue slides using a Chelex-100 resin extraction method. The extraction resulted in genomic DNA with an average yield of 800 ng from 20-µm specimen slides. The quality of DNA was verified using conventional PCR, demonstrating amplification of specific human gene fragments from 150 to 1100 base pairs in length (5).
Amplification of FFPE DNA
DNA extracted from FFPE tissue slides was amplified using GenomiPhi DNA Amplification Kit and DOP-PCR. Approximately 10–25 ng of FFPE-extracted DNA was used for each amplification.
When DOP-PCR amplification was carried out on FFPE- isolated DNA using three primers (DOP1, DOP2, and DOP3) as described in the Sanger protocol, only the DOP2 primer was successful in amplifying FFPE template DNA (Fig 1). The sizes of amplified DNA from DOP-PCR amplifications were small, with fragments ranging from 100 to 600 bases. The amplification was found to be dependent on the quality of FFPE DNA, resulting in lower yields than the GenomiPhi method (Table 1).
In contrast, GenomiPhi DNA Amplification Kit proved more consistent and reliable in amplifying FFPE-extracted DNA (Fig 2). The success of GenomiPhi amplifications was measured by the presence of a smear on the agarose gel indicat ing a high molecular weight (~20 kb) product. We have also verified the specificity of FFPE DNA amplification from GenomiPhi amplifications by performing conventional PCR on select human genes that yield products ranging from 150 to 1100 bp (data not shown).
Amplified DNA yield from FFPE tissue
FFPE-extracted DNA amplification using GenomiPhi DNA Amplification Kit consistently gave higher yields of product (~37 g per reaction), whereas amplification yield from the DOP-PCR method was lower and more variable (Table 1 and data not shown). The higher yields from amplified product show that this method is tolerant to degraded DNA samples such as those extracted from FFPE slides. The amplification consistency is caused by the strand-displacement activity of Phi29 DNA polymerase on starting template, which continues until all nucleotides in the reaction are exhausted.
Cy3 and Cy5 probe preparation
The sizes and yield of Cy3- and Cy5-labeled probes (analyzed on 2% agarose gel) were identical when both GenomiPhi and DOP-PCR amplified templates were used for labeling (Fig 3). The high molecular weight DNA made by GenomiPhi DNA Amplification Kit was found to have no impact on Cy3 and Cy5 dye incorporation; sizes and yields of probes were similar with or without Dpn II restriction enzyme treatment of amplified DNA (Fig 3). For array CGH analysis of FFPE-amplified DNA, two Cy3 and Cy5 labeling reactions were combined (10 µg total) for hybridization to give higher signal levels, which help in downstream data processing.
Hybridization of Cy3- and Cy5-labeled probes to BAC CGH arrays
The Cy3 and Cy5 labeled probes from both amplification techniques were hybridized separately on SpectralChip 2600, which is a high-density array CGH chip consti tuting approximately 2600 human BAC clones at 1-Mb resolution. A section of hybridization images from the BAC arrays is shown in Figure 4. The signal intensities from array features were quantitated to compare hybridization performance. The scatter plots of mean intensities from DOP-PCR and GenomiPhi amplifications showed hybridization characteristics to be similar between the two methods (Fig 5). Microarray hybridization data for mean intensities and background was compared between GenomiPhi and DOP-PCR to assess amplification performance between the two techniques. The results are summarized in Table 2.
amplifications. Right: Hybridization image from GenomiPhi amplifications. DOP-PCR and GenomiPhi amplifications showed hybridization characteristics to be similar between the two methods (Fig 5). Microarray hybridization data for mean intensities and background was compared between GenomiPhi and DOP-PCR to assess amplification performance between the two techniques. The results are summarized in Table 2.
The total signal from the arrays (5200 features, in duplicate) was similar between the two amplification methods. With GenomiPhi amplification products, nearly equivalent signal was seen from both the Cy3 and Cy5 dyes, indicating that DNA amplified with the Phi29 enzyme can be efficiently labeled and hybridized with both dyes. The mean signal intensities from DOP-PCR and GenomiPhi amplified samples were similar, with values averaged from both the Cy3 and Cy5 dyes of 315 and 381, respectively. The averaged signal-to-background ratios from DOP-PCR and GenomiPhi amplified DNA were 4.92 and 5.02, respectively. The BAC array hybridization results demonstrate that FFPE samples can be amplified with the GenomiPhi kit for array CGH profiling. Hybridization data from the low-density Constitutional Chip™ 1.0 BAC array (Spectral Genomics) w as also similar between the two methods (data not shown).
Because identical FFPE samples were amplified and labeled with Cy3 and Cy5 dyes, the log2 ratios from GenomiPhi amplified samples clustered close to the baseline level, i.e. 0 (data not shown). While no DNA copy-number changes were measured, previous data published using Phi29-based amplification has shown that GenomiPhi DNA Amplification Kit is highly sensitive in detecting monosomy and trisomy in array CGH (6, 7).
Our results showed that GenomiPhi DNA Amplification Kit can be used to amplify DNA extracted from FFPE tissue. Phi29-based whole-genome amplification, with its high processivity and strand-displacement activity, proved to be a robust method for amplifying DNA extracted from FFPE templates. The amplification yield from the GenomiPhi kit was higher than that from the DOP-PCR method. We also showed that FFPE tissues can be subjected to CGH analysis using samples amplified with GenomiPhi DNA Amplification Kit. Spectral Genomics human BAC arrays were successfully used to test the feasibility of array CGH using the GenomiPhi method. Comparison of hybridization results showed mean signal intensities and signal-to-background ratios of GenomiPhi amplified samples to be similar to those of DOP-PCR. Because GenomiPhi DNA Amplification Kit can amplify DNA from small quantities of template, this method can be used to study copy-number changes in FFPE samples on BAC CGH arrays.
Low yield from DNA extraction from FFPE tissue.
Use FFPE slides with at least 20 µm of tissue for DNA extraction. Complete the extraction method without a purification step because increased manipulation of the genomic DNA can result in loss of overall yield.
Low yield from GenomiPhi amplification of FFPE-extracted DNA.
Some extraction methods contain factors that will partially or completely inhibit DNA polymerases, in this case Phi29. Purification of the extracted DNA might improve results. Use of commercial kits such as Nucleon™ HT Genomic DNA Extraction Kit (GE Healthcare) will reduce or eliminate inhibitory factors and can be a useful substitute.
No signal observed from microarray hybridization.
Verify that Cy3 and Cy5 probe labeling was successful by analyzing sample on gel. Check for PCR amplification of select human genes (300–900 bp) from FFPE-extracted DNA and amplified DNA to make sure DNA is of high purity.
Log2 ratios distorted.
Use test and reference samples amplified under similar conditions. Make sure repetitive sequences are repressed by pre-hybridizing the arrays with human Cot-1 DNA™ (Invitrogen) and denatured herring sperm DNA.
1. Salman, M. et al. Will the new cytogenetics replace the old cytogenetics? Clin. Genet. 66, 265–275 (2004).
2. Kallioniemi, A. et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258, 818–821 (1992).
3. Fiegler, H. et al. DNA microarrays for comparative genomic hybridization based on DOP-PCR amplification of BAC and PAC clones. Genes Chromosomes Cancer 36, 361–374 (2003).
4. Mamone, J. T. A method for representatively amplifying genomic DNA. Genomic/Proteomic